Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
ABSTRACT
Existin 0 limit state formulations for tendon and steel riser bodies used in tension leg
platforms have been reviewed. These include AP1 Bull. 5C3 (for casinos ), API RP
2A LP, FD, BSI Sub-sea Pipeline, Danish Pipeline Code, DnV Submarine Pipeline
Standard and NPD Offshore Structures Regulations, as well as formulations
proposed by Johns and McConnell (Battelle), Jensen and Pedersen (Technical
University of Denmark), Loh (Exxon), Murphey and Lanoner (Shell), Tamano et
al. (Nippon Steel) and Winter et al. (TNO).
Ultimate strength formulations for tension, bending and pressure actin O singly or
in pairs of loads are compared. There are significant differences in the strength
formulae for pressure load, mainly due to the treatment of initial ovality; and for
bending capacity depending upon whether plastic or first yield bending capacity is
utilized. Some formulae for interaction between pressure and bending, as well as
pressure and tension, imply little mode interaction, while others, specially for low
diameter-to-thickness ratio, indicate strong interaction, i.e. almost a linear re-
lation.~rhip as in all formulae for tension combined with bending.
NOTATION
107
108 S. F. Estefen, T. Moan, S. Sa,vik, R. A. Zimmer
E Young's modulus
K Pipe longitudinal curvature
n Tvergaard hardening parameter
P External pressure
T Effective (applied) axial tension load
t Pipe thickness
INTRODUCTION
~0 -- [)max -- Drain
/)max "~-Drain (1)
EXTERNAL PRESSURE
TABLE 1
Ultimate Strength Predictions for Tubulars Under External Pressure (Initial Ovality 0.5%)
from BsI, rI'NO (Haagsma's equation Is) and Shell were close to the mean
predicted hoop stress values.
LONGITUDINAL BENDING
Apart from API-5C3 and Nippon Steel, which are aimed at casing tubes,
all other formulations mentioned are able to evaluate the bending capacity
of tubula~rs. However, there are two different approaches for treating the
ultimate bending strength depending on the tubular use as either platform
structural member or submarine pipeline. In the former, typically a
load-controlled resistance, the bending strength is defined in terms of
ultimate longitudinal stress, while in the latter, mainly due to the pipeline
installation, the definition is made based either on longitudinal strain or
curvature.
In the case of TLP tendon and riser bodies, the installation phase does
not necessarily imply longitudinal strain beyond the elastic regime. In
addition, environmental loads caused mainly by waves and currents can
precipitate low cycle fatigue in tubulars presenting elasto-plastic behav-
iour. Therefore, it is desirable in both tendon and riser applications to
avoid a design formulation which allows an ultimate plastic bending
behaviour. In the comparative study of the different limit state formula-
tions, an attempt is made to represent the bending strength as a function of
the ultimate longitudinal bending stress.
112 s. F. Estefen, T. Moan, S. Scevik, R. A. Zimmer
KD
e= 2 (2)
ff
=tr P~fnEe
ff p
+ l--n} for a>ao (3)
/
TABLE 2
Ultirnate Strength Predictions for Tubulars Under Longitudinal Bending
(Initial Ovality 0.5%)
it can be assumed that practicall~¢ all bending strength predictions are not
affected by the tubular initial ovality.
In all examples considered the maximum stress values were obtained
from TNO predictions and the minimum from either N P D or BSI. The
differences between maximum and minimum values were of the order of
45%, 40'% and 35% for D/t= 15, 25 and 35, respectively. In all examples
the Shell predictions were very close to the mean values.
AXIAL TENSION
All the formulations which take into account axial tension effects on
tubulars recommend the first yield to be assumed as the maximum value
for the axial tension stress. The only exception is for the DnV which
proposes a reduction factor to be applied to the yield stress for pipes with
D/t greanter than 20. For the geometrical range of interest of this study
(D/t= 12-35), it means a maximum reduction of less than 2%. As far as
limit state equations are concerned, it would be desirable to assume the
axial tension limit strength beyond the yield stress, but there is a
significant lack of experimental data on tubulars representative of casings
and pipelines to support a less conservative approach. Although several
experimental tests have been performed for casings under axial tension
loads, often they were interrupted when the respective longitudinal strain
reached the equivalent yield strain.
In case of formulations aimed at pipelines, the axial tension acting on
the pipe is normally considered as the applied axial tension less the
114 S. F. Estefen, T. Moan, S. St~vik, R. A. Zimmer
external pressure end cap effect. In this particular case the restriction in
terms of yield stress is related to the actual stress acting on the pipe wall.
0.9-
0.8-
g 0.7"
0.6-
'w~ 0.5 -
0.4-
............' ':re:i::,,
o 0.3- [ BATTELLE
0.2"
0.1"
0 -- i " ,£',..
o ola 0;4 o'.6 o'.a ~ 11a 1;,= l e
Long.bend.stress/yield stress
Fig. 1. Limit state formulations for external pressure combined with longitudinal bending
(D/t= 15, X-77, 6o =0.5%).
Limit state formulations for TLP tendon and steel riser bodies 115
0.7"
O.6~- ~ .."...-.! ~
~ 0.4- -"
~ 0.3-
a.
0
° 0.2
I- 1
0.1'
"~ u
0.35-
0.15
n i\ \ ":"...-:.-.ii_'..../\
IA~''"F° I ~(,"-? i
0 ~
0 0,2 0.4 0.6 0.8 1.2 1.4
Long,bend.stress/yieldstress
Fig. 3. Lirait state formulations for external pressure combined with longitudinal bending
(D/t=35, X-77, ~o =0.5%).
LRFD, DnV and N P D are not influenced by the initial ovality. However,
for a particular geometry, the respective shapes are practically unaffected.
The curve configurations are mostly affected by the effect of ovality on the
ultimate hoop stress under isolated external pressure. For D/t = 15 it can
be realized that API-LRFD proposes a significantly stronger interaction
between external pressure and longitudinal bending than the others. A
116 S. F. Estefen, T. Moan, S. Scevik, R. A. Ziramer
E X T E R N A L P R E S S U R E C O M B I N E D WITH
AXIAL TENSION
Eight formulations for external pressure combined with axial tension have
been analysed, API-5C3, API-LRFD, BSI, DnV, Nippon Steel, NPD,
Shell and TNO.
API-5C3 and Nippon Steel were based on empirical data for casings.
Collapse pressure expressions for API-5C3 were obtained from statistical
regression analysis of approximately 3000 experimental tests. The collapse
resistance in the presence of axial stress is calculated by modifying the
yield stress to an axial stress equivalent grade. However, API-5C3 for-
mulae are not valid for the yield strength of axial stress equivalent grades
less than 165 N/mm 2 (24000psi). Nippon Steel employed a similar ap-
proach based on 261 test data from commercial casings. An attempt was
made to include both ovality and residual stress in the collapse resistance
expressions.
Shell accounts for the axial tension loads by using the Von Mises yield
criterion to adjust the yield stress to be used in the collapse pressure
expression. N P D also takes advantage of Von Mises yield criterion
explicitly included in the interaction equation to treat this load combina-
tion. API-LRFD and TNO define pressure and hoop stress capacities,
respectively, in the presence of axial loads. BSI and DnV assume the
respective equations for pressure combined with axial tension in a similar
way to the previous proposition for pressure combined with longitudinal
bending.
As the formulations are aimed at either sub-sea pipelines or casings,
different effects due to external pressure should be taken into account. In
the case of pipelines, it is expected that during installation the pipe would
have an end cap which allows the effect of a compressive pressure loading.
Some collapse tests are performed in vessels which do not allow this end
cap effect to be considered. Axial tension equal to the end compressive
pressure load has to be applied to obtain a loading condition equivalent to
that simulated in this type of test. Therefore, the axial tension stress acting
Limit state formulations for TLP tendon and steel riser bodies 117
0.3.
, , ~ f",, ", f',~
0.2" : '. x ~
OA" i
0
0 0.'1 0.'2 0'.O 0'.4 O'.S 0'.6 0.7 0.'8 0.'9
A.)dal ten$.slress/yield stress
Fig. 4. Limit state formulations for external pressure c o m b i n e d with axial tension
(D/t= 15, X-77, 6o =0.5%).
on the pipe is obtained by deducting the end cap pressure effect from the
effective (applied) axial tension stress, as indicated in the following express-
ion:
T pD
o~=-7.a -- 4--t (4)
The different limit state equations for external pressure combined with
axial tension are shown in Figs 4-6 for tubulars with D/t = 15, 25 and 35,
steel grade X-77 and initial ovality of 0.5%. N P D and TNO are not
represented in these figures• TNO is similar in shape to DnV, which
predicts small mode interaction. The strongest interaction is given by the
casing-oriented formulations, API-5C3 and Nippon Steel. API-LRFD and
Shell curves present a intermediate behaviour as far as the interaction
between external pressure and axial tension modes is concerned. The
API-5C3 curve is interrupted before reaching the axial tension-dominated
region duie to its recommended range of validity, as mentioned above.
Four formulations are able to treat explicitly the combined effects of axial
tension and longitudinal bending, API-LRFD, BSI, DnV and NPD. All of
118 S. F. Estefen, T. Moan, S. Scevik, R. A. Zimmer
0.7
"-A ~'
~'~
0.35
~ 0.2-
~ 0.15-
~ 0.1"
0.05'
0
o o11 o12 o'.a o'.4 0'.5 0:6 o'.7 o.a o~9
Axialtens.stress/yieldstress
Fig. 6. Limit state formulations for external pressure combined with axial tension
(D/t =35, X-77, 6o=0"5%).
0.8
0.7" ", •
0.6
05
Fig. 7. Limit state formulations for axial tension combined with longitudinal bending
(D/t=25, X-77, (~o=ff5%).
LRFD and DnV are considerable less conservative than BSI and NPD.
The former formulations allow substantial plastic behaviour in the bend-
ing mode. For the axial tension mode, the DnV formula assumes a axial
tension resistance slightly smaller than the yield stress for the examples
with D/t= 25 and 35.
CONCLUDING REMARKS
Limit stalte equations able to treat tubulars with geometry and material
properties typical of TLP tendon and steel riser bodies have been
reviewed. They were obtained from design codes and published papers
aimed at design recommendations for casing tubes, submarine pipelines
and platform structural members. Loading conditions represented by
external pressure, longitudinal bending and axial tension, acting alone or
combined in pairs, have been considered.
Experimental results for tubulars with relevant material, geometrical
properties; and load conditions have been reviewed in order to evaluate the
reliability of the formulations. Correlation between strength predictions
and test data is presented elsewhere.19
Owing to the different approaches associated with the respective formu-
lations as far as bending strength was concerned, both strain and curva-
ture at collapse were transformed to longitudinal stress by using the
Tvergaard expression in the plastic range.
120 S. F. Estefen, T. Moan, S. Scevik, R. A. Zimmer
ACKNOWLEDGEMENTS
REFERENCES